Tip:
Highlight text to annotate it
X
This is a live event where we're going to discuss topics and questions about our second
Analytics Academy course, Google Analytics Platform Principles. We're really excited
about this course, and we have over 100,000 students registered all over the globe. This
course builds off of some of the ideas that we talked about in our first course, Digital
Analytics Fundamentals. Today, I'm excited to be joined by Sagnik Nandy, who's a Principal
Engineer on the Google Analytics team. Welcome!
Well, very very happy to be here.
So why don't you tell us a little bit about what you do on the team.
Ah, so, my team manages everything to do with data in Google Analytics, and we have a lot
of data! So that ranges from collecting the data, storing the data, making it query-able,
so adding additional constructs and tools on top of the data -- so segmentation, filtering
-- all of those. My team also manages the Real-Time system, the API, the Developer Relations
side of it. So anything to do with ingesting data, saving data, and making data completely
accessible to end-users, that's what kind of we do -- that's our business.
So, that's the whole product! It sounds like everything!
A fair bit of the product -- there are lots of really bright engineers doing other things.
Awesome. Great. So what we're going to talk about today is we're going to spend a little
bit of time talking about the material in the course, the Google Analytics Platform
[Principles], and then we're going to get to some questions that were submitted by you,
our students. So this course, Google Analytics Platform Principles, we spend a lot of time
talking about the four parts of the platform -- collection, processing, configuration and
reporting. So, that's obviously something you know a lot about. So, in your own words,
can you tell us a little bit about why it's so important to really understand the platform?
So, I think the platform is basically the underlying tool that makes everything -- all
your analysis -- possible. So, I think there are two big benefits of understanding the
platform well. One is it gives you a better understanding of the reports you're looking
at. And this is often non-intuitive, but the same report can be interpreted differently
by different people, and that might range from an accurate interpretation to a completely
incorrect one. So, I think understanding the platform, that you understand what you're
looking at, how to look at the data, what insights you can gain from it. A good example
would be something like bounce rate. Now bounce rate is very very critical for measuring engagement,
and let's take the example of a blog. For a blog, a lot of times people just come to
a single page, and they spend their entire time looking at that page. Even though they're
engaged, it might just be one page view, and a bounce. So sometimes what people do is they
would send us additional Event pings to make sure it's not a bounce, but in order to do
that you have to make sure those events are marked as interactive Events and not non-interactive
events. So understanding these nuances can really help you understand the report you're
looking at. If something looks bad, is it really bad? If something looks good, is it
really good? Et cetera. I think the platform is that it makes you realize what you can
do with the system. So a good example analogy would be, imagine if you have a smartphone.
You can just use it for making phone calls, but there's so much more you can do with it.
In order to understand that, you need to understand what the various options are. So I think of
the platform that way. You can just look at reports, but there's so much more -- you can
create your own reports, you can segment in certain ways, you can set up your own filters
-- all of these things. Once you understand the platform, it allows you to do so much
more. So to kind of summarize, it helps you understand what you're looking at much better,
and B, it helps you look at more things -- look at things more differently -- et cetera.
Yeah, and I think you brought up a really good point of, kind of, being able to grow
your Analytics, right? Because business is changing rapidly in the digital space with
the onset of mobile and all of the different touch points we have with consumers, really
you don't need a reporting tool, you need a platform that's going to help you gather
all the data that you need from all of your digital touch points, and then synthesize
it into actionable reports. And if you don't understand all of the different parts to that
platform, then you can't evolve or expand your Analytics practice as the business may
grow and move into different digital places.
Completely agreed. My personal hope is that every person using our product makes it a
slightly different product based on how they use it. And I feel the platform is what allows
them to make that happen.
That's great. So one other thing that's really important and something we talked about extensively
in our first course actually isn't really the platform, but it's the business, right,
and understanding what's happening from a business perspective. And we talked a lot
about creating a measurement plan for your business. Now the measurement plan is one
piece, the platform's another piece, and then you kind of combine them and you can create
an implementation plan, right, that tells you how to implement Analytics. So maybe you
could tell us a little bit about once you have a measurement plan, how do you kind of
translate that based on your knowledge of the platform into implementation steps?
And again, as you say, there are three key pieces here: your business plan -- so hopefully
if you're going through the pains of tracking your business, doing analysis, there are certain
goals you want to achieve, so figure out what is it you want from your Analytics tool; and
then take a look at the platform and see what does the platform provide you. We were just
talking sometime before this and I mentioned that you're only as powerful as the tools
you have access to, so your platform is the tool set here. So it's like if you were trying
to fly and you had access to an airplane versus you had access to a car, it would be two very
different problems. So figure out what your platform allows you to do, map your business
problems to those tools, and then see what data do you need to feed in and take out in
order to achieve those objectives. I mentally think of them as the three A's. I keep saying
them. So you have Accumulate -- so you want to bring in data, you want to Analyze, and
then you want to take Action. All of these should map back to your business. There are
lots of things. A good example would be conversions. So Analytics as a platform allows you to measure
different kinds of conversions -- we have goals, we have transactions -- so you should
first ask what is it that you're trying to achieve for your business. I used the example
of a blog. Maybe for a blog their goal is just to have user engagement -- they want
to get more users. So in that case, that's what you should try to measure. Maybe for
a blog they show ads and getting clicks on those ads is their main business, so that
becomes your objective. For a retailer they're selling things, so then they would heavily
use our eCommerce offering as opposed to our Goals offering, because there are special
advantages for using our eCommerce offering for a retailer. So you need to understand
the platform, the data model. We have specific analysis which is only available, like for
example, for conversions, we have a very powerful attribution and funneling support. So by mapping
things to conversions you can get that kind of analysis. So map your business problems
to the tools and that way you'll be able to leverage maximum benefit out of it.
Yeah, that's an awesome answer. And I know people always ask, you know, if you had to
pick three things in Google Analytics, what are the basic things that have to get set
up. We've always talked about, like you just mentioned, setting up conversions. I think
another feature that's really important is the campaign tracking, right, because that
helps you identify where people are coming from. So conversions help you measure, are
people doing what I want. The campaign tracking helps measure are they coming from my different
marketing activities, like email. What are a couple of other features that you usually
like to talk about when getting your implementation set up the right way?
Right, I think you covered on some of them. The three most important ones in my mind is
conversion -- be it Goals or Ecommerce Tracking. Filters would be another big one.
Filters is a great one.
And the third one would be Channel Grouping, which kind of falls under the bucket of campaign
tracking. I'm surprised at how many people don't do any of these. So one tip for everyone
watching this: if you have just these three things implemented, you have already created
a huge strategic advantage for yourself as a marketer or as a practitioner. You'll be
surprised at how many people overlook the power of these tools. So to get back to the
three, so for conversions, as I mentioned, you probably have a certain goal and objective
which you're trying to optimize. And so make sure you have your conversion set up. Over
the years we've added support for all kinds of conversions, be it engagement, be it actual
tracking URLs, tracking Events, tracking Ecommerce for retailers. So there's a whole bunch of
conversion support and we keep adding more to that tool set. Filters. It's highly unlikely
that the data you're sending us is completely purse and pristine. So it's one of your biggest
responsibilities to get good insights out of data is to make sure it's good data going
in. Garbage in, garbage out.
Garbage in, garbage out -- it's what we always say.
A good example would be, imagine if you have an internal test team that hammers your site
to test how powerful it is, how much load it can withstand. You probably want to drop
that traffic before you show your overall performance report to your CMO. So filters
allow you to drop your internal traffic, allow you to segment by groups of people. So if
you want to set up a set of reports for your US visitors versus your visitors from Europe,
it allows you to do that, so please take a look at Filters. It helps you both enhance
the data as well as clean up the data. And thirdly, Channel Grouping, which leads to
campaign tracking. We did an entire revamp of this a few months back, and the reception
has been very powerful. We want to give you the tool so that you can map your acquisition
the same way you think of your business. A good example would be, imagine if you're paying
a third party site to refer traffic to your site. Through Google Analytics that would
like like referral, because we don't know about this arrangement you had. But you know
that counts as paid traffic, and you want to let us know that. So with Channel Grouping,
we allow you tools to say, you know, even though this looks like plain referral traffic,
this is actually affiliate tracking or paid traffic. Or something might be social to you.
Like you might be running a separate sub-site which allows people to share their experiences.
To us that's just another page, but to you that's social. So for you to be able to categorize
your traffic in different buckets, so that you can then get the best from an acquisition
point of view, is very important.
Yeah, and I think, you know, kind of tying this back to the whole platform view, a lot
of these features, they build on top of each other. Like, I love how you talked about Filters
right there, because kind of Filters maybe would be considered maybe like a foundational
thing, because it really cleans up your data. Like it purifies it so you don't have all
that garbage, right. And then Goals are like the next level, because you absolutely need
to measure conversions, and then as you keep leveling these things on, like, the actionability
and the usability of the data just increases almost exponentially. And I think a really
great example of that is we have this amazing feature called Attribution Modeling, right,
and the whole Multi-Channel Funnels reports, which are phenomenal, right. Like you get
to see this multi-session behavior and how your different marketing activities drive
conversions, but you can't use that feature unless you've got Goals set up, unless you're
doing campaign tracking, unless you're kind of customizing your Channel Groupings. And
so, like, those three things work together to then provide this amazing feature, to me,
which is really one of the most powerful parts of Analytics. So you gotta get all those things
right.
I completely agree.
Cool, so when we talk about the Google Analytics platform, and specifically in this course,
we talked about the current platform. And some of our students, and some of the folks
out there in internet-land might know this as Universal Analytics. It's something we've
been talking about for a little bit. Can you tell us a little big about how Universal Analytics
and the current platform differs from the previous version of Google Analytics?
Completely. So first and foremost, I'm super super excited about Universal Analytics.
This is Sagnik being super excited by the way, just so everyone knows. This is super
excited Sagnik.
Super excited about it!
You should see kind of calm and passive Sagnik.
[LAUGHTER]
We've invested so much engineering resource and from a platform development point-of-view,
in Universal Analytics, I can on the record and say Universal Analytics is going to be
the future of analytics. Like our entire future strategy is around it. So if you are sticking
with the old version, you need to change, because others will, like, rocket past you.
So please please please consider Universal Analytics. It's been a little more than a
year since it's been out, and adoption has been phenomenal. But Universal Analytics allows
a whole bunch of things. To me, the three key biggest things it allows is A) it opens
up the measurement protocol, and that's been huge. What it allows people to do is take
Analytics beyond web analytics to just analytics. So you no longer just have to track your website.
You can tie it back to your business, you can tie it back to your store, you can tie
it back to your user data. So it opens up the protocol and says you can send us data
from anywhere and you can send us any kind of data. So we've seen people do fascinating
things. We've seen people track their coffee machines, we've seen people track their point-of-sale
systems. By opening up the protocol it really opens up the platform.
And the Measurement Protocol, that's the "hit" -- that's literally the transmission mechanism
of how you send us a "hit" of data.
Exactly. And that is no longer tied to your JavaScript. It's no longer tied to three or
four cookies sitting in your browser. By freeing up that, we've essentially empowered the entire
user base to do a wide range of things. The second big thing that Universal Analytics
allows is that it allows you to set your own User ID. Increasingly, businesses are trying
to look at users, and not sessions or cookies, by giving you the ability to track a user
in the same way. Your business thinks of a user, so to you a user might be a sign-up
ID for your business. To you a user might be a loyalty card number. We allow you to
map your users in the same way you think of it. And there are a whole bunch of things
you can do with that. You can do cross-device tracking, you can do cross-device attribution,
you can actually -- there's a whole gamut of things that you can do. So that's the second
big thing. The third thing is that we've really opened up the ability to send us custom data,
be it through Custom Dimensions and Metrics, be it through Dimension Widening. With Universal
Analytics we really allow you to map your data model to what you think of. So for example,
if demographic data is a critical part of your business -- so imagine if for every person
visiting your site, you somehow know if they fall under paid membership or free membership
-- you should be able to send us that kind of data, even though Analytics doesn't know
that, you should be able to augment the data model with these two attributes -- membership
type and maybe something else. And that what that becomes a first-class entity in your
data, and now you can segment all your reports. So if you're looking at your entire sales,
you can say I want to see what fraction of my sales comes from my premium users versus
my "freemium" users, et cetera. So it really allows you to map the data in your own way
and really opens up the platform and the power of it.
Yeah, and I think you brought up a really really good point there. You know, a lot of
the stuff that you just mentioned is all about the data, right. And so if we have Analytics,
and now it can ingest lots of different types of data, while the tools within Analytics
can still function in their same normal way on that data. And so now, if the data's coming
from kiosks or if the data is coming from set-top boxes, you can still do segmentation,
you can create Custom Reports, you can create custom Dashboards. You can use all of those
tools with that same pool of data, which is amazing.
I mean, I couldn't have said it better. One of the things we focused on in the last few
years is to separate the analysis tools from the data coming in, as opposed to having them
strongly tied, which as you mentioned, allows people to send any kind of data, and then
apply the same sophisticated tools that we've built on top of it. So to take your kiosk
example, if a particular user uses both a kiosk and later a site and you can track them,
now you can do segmentation based on them, so you can ask questions like, "how many people
did A followed by B?" So you can see how many people were acquired via a kiosk or were actually
a website user that then laster used your kiosk somewhere else. So, we've really separated
the two and in the process made the overall platform more powerful.
So one of the features that we got a few questions about, and one that I'd like to explore a
little bit more closely with you is the User ID feature. I think this is the one that is
probably going to be the most powerful for businesses, because it does exactly what you
said -- it allows businesses to use their identifier in the Analytics data. And so if
an airline, that might be a frequent flier number. And so when I'm logged into a kiosk,
when I'm logged into my app on a mobile device, when I'm logged into the website, that ID
-- that frequent flier ID -- can be used, kind of as the glue to hold those sessions
together. Can you tell us a little bit more maybe about, like how a business would go
about literally doing that -- adding that ID?
So there are multiple ways of doing it, and again, I think the true strength of the platform
is there is no one solution, so everybody can innovate with the platform. We've seen
lots of interesting things already being done. So to take the offline conversion example
that you kind of suggested -- let's take the airline example. So imagine you come to a
site. So at that point your JavaScript has access to the client-side cookie value, and
then you go to the site and you enter your loyalty card. So what the airline company
can do is actually log that information on their side, so they now have a mapping of
the client-side cookie value, which is what Analytics was receiving from the JavaScript,
and the corresponding loyalty card ID. And later someone actually goes to the airline
kiosk and actually does some action using just their loyalty card ID. But now the airline
actually had access to the corresponding client ID, and because we've opened up the measurement
protocol they can send us a hit saying, by the way, this client ID then later checked
into this physical location. And this is only possible because the kiosk is no longer reliant
on cookies. We allow you to set up the user ID any which way. Eventually with time -- and
we are already seeing this -- people will build these kind of ID controlling services,
et cetera. So some people have done interesting work around that. But it was literally as
simple for an airline company to map their client ID with their loyalty card ID and then
sending us hits mapping -- doing a reverse mapping. So it's very very powerful.
Yeah, and so that's the key, you know, businesses need to have that ID, right. And obviously
if you don't, that's fine. You can still use Analytics. But you just don't get some of
the advanced features like the cross-device measurement and things like that.
And again, I think the critical thing to keep in mind is that it need not be an exact science.
If you have a single logged-in ID, that's great. On the other hand, if you only have
a partial subset of users who log in, that's great, too. I might only get logged in information
from ten percent of my users, but I think I can get a lot of insights from that ten
percent and extrapolate that. Sometimes my users might be partially logged in. Sometimes
there might be incentives to allow a user -- a survey might be a good tool -- to allow
users to give in some kind of information which might not be personally identifiable,
but is still better than just a first-party browser-dependent cookie. So ask yourself
what kind of user-level information your business has access to. And on our side we've taken
amazing amount of effort to make sure things are done in a privacy-sensitive way -- that
we are respectful of the end-user privacy, constraints, et cetera. So really asking the
question, "how can my business leverage this solution?" I'm sure there's some way ranging
from complete logged in experience to absolutely anonymous.
And again, the value of a platform is you can work within it. It gives you lots of different
-- I kind of like to think of it as duct tape. You can duct tape up a solution that's best
for you in your current environment, and, you know, as you grow you can undo a little
bit and redo it again. So obviously -- excuse me -- a lot of people are excited about Universal.
I've been really excited about it. I feel like we've been talking about it for a long
time. But there are a couple of parts that are not at parity with the existing GA. So
can you tell us a little bit about, well, when can we expect this, and then, you know,
the future of Universal.
Why would you say that?
[LAUGHTER]
No we are very aware. It's been, as I mentioned, just about a year post-launch. We've reached
parity for a whole bunch of features in the last few months. I'm not going to give an
exact date, but what I am going to say is very very soon. Very very very soon. I can
go on for the rest of this Hangout just saying very very very soon. You'll have full feature
parity plus a whole bunch of extra new features which will be Universal Analytics only. So
as I mentioned in the beginning that, in my mind, there will be no reason for people to
stick with the old platform. Universal Analytics is like the new car that runs on electric
and doesn't need charging.
[LAUGHTER]
So I think there's also something that's really important here to understand is that if you
haven't started thinking about Universal Analytics you should probably do that. Now obviously
it's not like Universal's going to come and the old version is going to go away. Everything's
going to work fine. But you should really have a plan for your business to change from
the standard version of Google Analytics to Universal Analytics, because it's, you know
-- you've gotta make some coding changes. You need to do a small change to your data,
which actually isn't a lot of work. So having a plan in place is critical. And you might
even want to consider moving to tag management, right, if you're going to upgrade. Google
Tag Manager will support Universal Analytics. And if you don't use tag management this is
a fantastic time to kind of reevaluate that, because tag management will make everything
easier, not just Universal Analytics.
Completely agree. Like we already have first-class support for Universal Analytics in Tag Manager.
So it really facilitates. We have a section in our configuration now that allows you to
migrate to Universal Analytics. One of the big things that we added was you don't lose
your past data gathered from non-Universal Analytics days when you do a switchover. And
a big part of our team what they're looking at is to make this transition very very smooth
so that people don't feel any kind of hiccups.
Alright, well, let's move onto some questions. We could have done just all questions for
this time, but we wanted to talk a little bit about the platform. So one of the first
questions comes from Ben, and Ben asked us a question via the forum. And he asked a question
specifically about Lesson 2.2, which is, we talk about collection in Lesson 2. And his
question has to do with the User ID feature. So Ben asks, can anyone share more detail
on how to create your own identifier and maybe some of the do's and don'ts associated with
using an identifier. My goal is to tie offline transactions that happen after a site visit
to site activity reported within Google Analytics.
So Ben -- what can I say -- been there done that. We just kind of attacked a similar problem.
So for example, if you have some kind of user identifier, be it a loyalty card, be it something,
you know, about the user, at the site level, then you can map your client-ID back to that
identifier. And hopefully when the offline transactions or point-of-sale interactions
happen, it happens with that user identifier so you can then map it back and send us IDs.
As I mentioned, people have user ID brokerage services, et cetera. The important things
to keep in mind is, as I said, data need not be 100 percent -- of 100 percent coverage.
The other thing is, as I mentioned, we are very privacy sensitive, so we don't want you
to send us personally identifiable information, which is globally personally identifiable.
It's okay if you know something about a user but you can map it to something which then
obfuscates the user to us. So a classic example would be, imagine a user enters their name
in your system, we would not like you to send us their name, but then you can map that name
to some ID and then send us that ID. So that way we know with that ID that this is the
same user across multiple touch points, but then you can take that information out and
then map it back to your user base, et cetera.
Yeah, I think that's a great point. And, you know, it's some type of string -- some group
of characters -- that only you would understand. So our next question comes from Clark, and
Clark asked us a question via our Google+ community for students. And this also has
to do with collection. The question is, is it possible to use Google Analytics for non-Analytics
data. For example, I'd like to collect metrics for a Google Spreadsheet. How can I use Google
Analytics for things other than a website?
Great question. And that's our big hope. I've said this before -- it's not Google web analytics.
It's Google Analytics. So we really encourage and want people to track their entire business,
both on site as well as off site, using Google Analytics. And the Measurement Protocol is
what facilitates this. I gave examples of people tracking their coffee machine, their
point-of-sale systems. And there's been really innovative stuff, like, just search for Universal
Analytics and you'll see tons of crazy slash innovative solutions people have built. But
I've seen users who actually have nothing which is online and still track using Universal
Analytics. So they have complete offline businesses. They send pings to us so that they can use
our segmentation tool, our analysis tools, our visualization tools, but their entire
business is offline. With the Measurement Protocol, especially with Custom Dimensions
and Metrics, you can completely customize the data model, which I think is very powerful.
Yeah, and so, again, that Measurement Protocol is kind of like the core. It's how data is
sent to Analytics. And so we actually use the Measurement Protocol when we build the
JavaScript that people might use on their websites. It's also used in our SDKs for mobile.
So if you want to track your iOS app or your Android app, there's a Google Analytics SDK
that you can add in, and it, kind of, makes it easy to send the data, right. And if you're
doing something that's totally different, like we've talked a lot about like a kiosk,
you would actually embed the Measurement Protocol into the code that's running on that kiosk.
So you would have to find your nerd -- get to your developer -- and say okay when someone
touches this button I want you to send this hit of data off to Analytics.
Again, continuing with my series of bad analogies. I think of earlier we were only in the business
of building houses. Now we still build great houses, but we also give you the the building
blocks -- the bricks, the mortar, the paint and everything. So if you like the house we
built, you can stay in it. If not, you can build your better house with our tools.
And actually, Clark, just to answer your specific question about Google Spreadsheets, you actually
can't track Google Spreadsheets or Google Docs with Google Analytics, because there's
no way to add the JavaScript into those things. So unfortunately can't track those, but you
can track all sorts of other things, like dog collars, coffee machines, stuff that you'd
want to track all the time.
Okay, let's move on to another question. This question comes from Trey and it's about configuration
settings. Could you provide more details on what configuration settings I should use,
or can you tell me about which ones are important and often get overlooked?
I think we covered this earlier, but the three key ones I would mention again are Filters,
which help you clean up your data, conversions, be it Goals or be it Ecommerce transactions,
which then let you specify -- almost classify your traffic as either valuable versus the
rest of your traffic. And then finally Channel Group, which lets you kind of map your acquisition
to your business needs.
And I'll throw in a plug for, you know, I think often ones that get overlooked are ones
that may seem a little bit more involved in terms of setting up. I am a huge fan of Custom
Dimensions. I absolutely feel like that's one of the most important parts of Analytics
because you're sending custom data, right. And it aligns Analytics more closely with
what you're trying to do. So I love Custom Dimensions. And I'm a huge fan of Events.
Event Tracking let's people get more granular information. I've spent a lot of time helping
people figure out how to track like scrolling, so you can see people reading on pages. So
they're a little big more involved, but they increase the data and the more actionable
data, in my opinion.
Completely. And another plug for Google Tag Manager. if you're using Event Tracking, Google
Tag Manager is actually taking this whole configuration driven approach off Google Analytics
and extending it to Events. So they have very good first-class support for tracking Events
using Google Tag Manager. The other thing I would mention is there are a lot of advertisers
in Google Analytics user base. Consider linking your accounts. That's very very important.
We are surprised how often the same user has an AdWords account and an Analytics account,
is trying to track the same thing, but they won't have linked them together. Please do
that, because we've built a lot of tools on both sides to leverage this overall ecosystem
of advertising and onsite or offsite analysis. So by linking the accounts you will unleash
a whole range of tools and analysis techniques.
Yeah, that's a great one -- the account linking. And actually we have a question from Lucy
specifically about account linking. And it goes something like this: when linking the
AdWords and Webmaster Tools accounts to Google Analytics, is the data actually imported so
that if the accounts were ever unlinked would the historical data still exist in the Google
Analytics reports?
So great question. This one is a little more nuanced. So Webmaster Tools and AdWords, there
are two separate answers. For Webmaster Tools it is truly an on-the-fly importing. So if
you unlink the data the data is not persistent on our side. So as you unlink the data completely
goes away. The good thing is that when you link back the data completely comes back.
AdWords is a little more nuanced because what happens there is you do send us some information
through the hits. So you would have seen the "gclid" parameter in the URL, which sends
us that a particular click happened which we can then widen -- bring in more information
from AdWords. So the key thing to keep in mind is the widening happens on-the-fly, but
the fact that the click happened and you send us that information is persisted. So if you
unlink an AdWords account you'll miss a lot of the widened information. Like you'll see
the impression count, which we import on-the-fly, go away. A lot of the campaign data and the
information associated with it go away. But basic things like the fact that this happened
because of a Google click will still stay. That kind of information.
Yeah, so you'll still be able to see here's the number of conversions because of Google
AdWords, here's the number of transactions because of Google AdWords, but you won't be
able to see things like Ad Slot or the cost data.
Exactly.
Ok, cool. That's a good one. Great. That was a really cool question, and it's cool to dig
a little bit deeper into some of the inner workings. So we had another question. This
question is about Dimension Widening, which is a fairly new feature that we have in Analytics.
And the question is, can dimension widening be applied to the Event dimensions, like Category,
Action and Label? Or how about the campaign dimensions, like Campaign Name? And I just
want to remind people, Dimension Widening is how you can import data into Analytics.
And the fundamental way it works is you define a piece of data in Analytics and you define
the same piece of data in like a file, and that becomes the key that joins up and you
can map the data together. And so the question is here about can Events or Campaigns be used
as that key to join the data.
So not right now. We've kept the set of fields you can widen from limited, but very soon
again. Our objective, I can share -- our overall goal is to make every dimension both widen-able
as well as to be able to use it as the source of widening. So that's kind of what we are
aiming towards. Obviously there will be constraints around scope, et cetera. So for example, if
you are widening on the visitor-level information, there might be restrictions on what other
dimension you can widen it to, et cetera. But our goal is, within the data model, to
allow everything for both widening and to be used as a source for widening.
Wow. That's great. So a little coming attraction.
So stay tuned.
Stay tuned. Awesome. Cool. Alright, so I think we have time for one more question. I've got
one here from Sean. And this has to do with reporting -- obviously another thing we covered
a lot in the course. And it has to do with the combinations of dimensions and metrics,
which we talked about specifically. And so Sean actually has a question about creating
a specific report, and he wants to create a report that shows the number of Unique Visitors
for a page on my site. And I think that gets into the general idea of the scope of data,
right, and so you have like user- and session-level scope and hit-level scope. So why don't we
talk a little bit about that and how Sean might be able to create a view of this data.
Right. So scope is a very very important concept. One which we hadn't exposed very nicely in
the past. We are trying to fix that. Our new metadata model is highly scope specific, and
expect changes to that soon. But we think of data in terms of three scopes largely:
user, session and hits. So going forward, we'll try to group everything in terms of
these, and then we'll also set up very well established rules about what is allowed with
what other scope. Because sometimes, for example, people get confused. A good example would
be if you're looking at something which is a hit-level dimension with a session-level
scope, it can be misleading. An an example would be if you're just looking at URLs versus
visits. So visit is at the session-level as a metric. URL is at the hit-level. So a lot
of URLs will return zero for sessions. I think the question people are trying to ask is,
"in how many sessions did this URL appear?" I think the answer we are giving is, "how
many times was this URL the start of a session?" So these confusions happen and we've gotten
a lot of requests. So we're going to make scope more first-class in our metadata. So
using the API, and even the front-end, you will be able to deal with scope better. But
to get to the specific question, I think it's okay to propagate visitor-level metrics -- user-level
metrics -- in this case, users, with hit-level dimensions, and we take care of it. So if
you ask for the dimension URL or Page Title, depending on which is a better map of what
you call "page," and request for the metric Visitors, you will see Unique Visitors. You
will see the report that -- I think Sean?
Sean was requesting. So, there you go Sean.
Awesome. So that wraps up the questions. And I want to thank you so much for coming. This
was a lot of fun.
Oh thank you. You guys are doing an amazing job. Seriously, you're literally taking Analytics
to so many more people. We are super thrilled, and thank you so much for having me over.
Awesome. And obviously thanks to the students! A few things to tell you guys as we wrap up
here. First and foremost, don't forget that this course, Google Analytics Platform Principles
is open through Thursday, March 27. So that means if you would like to earn a certificate
in this course, you need to complete all the lessons, all the assessments and the final
assessment by Thursday, March 27. Now, after Thursday, March 27 all of this material will
still be available. You'll still be able to go to the website and still be able to watch
the videos, but the big difference is that the course certificate will no longer be available.
So you won't be able to earn a certificate after March 27. So if you want to do that,
March 27 -- that's the deadline. You have another week and a day. Also wanted to mention
a big note of thanks to the entire team that is bringing this Hangout to you and to this
entire course. Christina, Gary, Chris -- it's a big effort. Catherine was a huge help. I
really appreciate it. So thanks to those guys. And stay tuned. We're working on a lot more
courses. We've got a lot more material planned for you guys, and we're super excited to launch
our next course very very soon. So that's going to wrap it up for today. Thanks again
for joining us everybody, and have a wonderful day!